A method and system for identifying and processing accumulated residue on a conveyor belt of a heading machine

Abstract

The present application relates to a kind of tunneling machine conveyor belt accumulated residue identification processing method and system, under the tunneling state of tunneling machine, the image of target area of the position where conveyor belt of tunneling machine discharge port below falls residue is collected;Sparse optical flow of target area is obtained, and the proportion of optical flow in target area is calculated as flow rate;If flow rate is less than set value, it is judged that conveyor belt accumulates residue, and executes accumulated residue processing.The present application collects the monitoring image of target area below discharge port, identifies the part of image moving by optical flow method, and further calculates the proportion of moving part (optical flow area) in target area, if moving part proportion is less than set value, then consider that conveyor belt appears accumulated residue at discharge port, realizes the unmanned, automatic identification of accumulated residue, improves the intelligent degree of tunneling machine discharge system and tunneling machine itself.

Description

TECHNICAL FIELD

[0001] The present application relates to a kind of tunneling machine conveyor belt accumulated residue identification processing method and system, belong to tunnel shield construction accumulated residue removal technical field, can be used as the powerful supplement of the automatic adjustment of soil bin pressure and the automatic adjustment of belt when shield unmanned tunneling. BACKGROUND

[0002] The belt conveyor system and screw conveyor (hereinafter referred to as "screw machine") system in the earth pressure balance shield machine are one of the important subsystems of the shield machine, the screw machine system realizes the transportation of the excavated soil to the belt conveyor system after the cutterhead excavates the soil, the belt conveyor system realizes the continuous transportation of the excavated soil to the muck car after the screw machine discharges the excavated soil, and the screw machine system and the belt conveyor system play an important role in the conversion of the excavated soil from the cutterhead to the screw machine, the belt and the muck car. At the same time, the speed of the screw machine system is one of the most important factors affecting the size of the soil bin pressure. The state of the excavated soil is the key to whether the belt conveyor system can successfully transport the excavated soil. If the water content in the excavated soil is large, it is easy to cause the belt conveyor system to fail to transport the excavated soil, so that the excavated soil continuously accumulates at the conversion place of the screw machine and the belt, causing the initial segment of the belt to have too much pressure and unable to realize the continuous transportation of the excavated soil, thereby affecting the tunneling efficiency of the shield machine.

[0003] Similar problems also exist for other types of tunneling machines, such as full-face hard rock tunnel boring machines (hereinafter referred to as TBM), the excavated soil is poured onto the belt conveyor by the excavated soil discharge port of the excavated soil discharge system, and the excavated soil is discharged by the belt conveyor. If the excavated soil accumulates at the position where the belt conveyor falls, it will affect the normal tunneling of the TBM.

[0004] The existing identification and processing means for preventing accumulation of excavated soil at the conversion place of the belt conveyor mainly rely on installing a camera at the conversion place of the belt, and the main driver observes the discharge of the excavated soil by the camera in real time, including the water content, whether it is gushing, and whether it is accumulating at the conversion place of the belt, and then adjusts, for example, the water quantity of the cutterhead of the shield machine or adjusts the speed of the screw machine, to ensure that the screw machine discharges the excavated soil smoothly and the belt conveyor transports the excavated soil smoothly.

[0005] However, on the one hand, when the water content of the excavated soil is large, the discharge of the excavated soil by the screw machine may gush, which may contaminate the camera and limit the view of the main driver at the conversion place of the belt.

[0006] On the other hand, the main driver needs to constantly pay attention to the accumulation of excavated soil at the belt conveyor through the camera and the display, which causes distraction of attention during tunneling and is not conducive to tunneling safety.

[0007] At the same time, with the improvement of intelligent level, autonomous tunneling of the tunneling machine is the future development trend, and the participation of the main driver in the conventional tunneling needs to be eliminated. Therefore, an automatic identification and processing method for accumulation of excavated soil on the belt conveyor during tunneling is urgently needed. SUMMARY

[0008] The application aims to provide a tunneling machine conveyor belt slag accumulation identification processing method and system to solve the problem of dispersing the attention of the tunneling machine driver in the prior art, which is not conducive to improving the intelligent level of the tunneling machine.

[0009] To achieve the above-mentioned purpose, the application provides the following solutions.

[0010] The tunneling machine conveyor belt slag accumulation identification processing method of the application collects images of a target area at a slag falling position of a conveyor belt below a slag discharge port of a tunneling machine slag discharge system in a tunneling state of the tunneling machine; obtains sparse optical flow of the target area, and calculates a proportion of the optical flow in the target area as a flow rate; if the flow rate is less than a set value, it is judged that the conveyor belt is accumulated with slag, and slag accumulation processing is performed.

[0011] The application collects monitoring images of a target area below the slag discharge port of the slag discharge system, identifies the moving part in the image through the optical flow method, and further calculates the proportion of the moving part (optical flow area) in the target area, if the proportion of the moving part is less than the set value, it is considered that the slag falling position of the conveyor belt is accumulated with slag, realizing the unmanned and automatic identification of slag accumulation, and improving the intelligent level of the tunneling machine.

[0012] The set value can be calibrated according to the actual situation on site, considering that the moving area in the target area is mainly the surface of the conveyor belt and the slag soil on the surface, so the proportion of the surface of the conveyor belt in the monitoring image can be taken as the set value; or for example, the minimum critical value of the proportion of the moving part (optical flow area) in the target area when the conveyor belt normally conveys slag soil can be found as the set value.

[0013] In the prior art, the slag accumulation at the slag falling position of the belt conveyor is collected and transmitted by the camera to collect the images of the target position, and is judged by the tunneling machine driver by artificial naked eye, the application can be improved on the basis of the prior art, and the original camera is used to collect the monitoring images, and the image processing device is added.

[0014] The tunneling machine conveyor belt slag accumulation identification processing method of the application solves the problem of dispersing attention by manual observation in the prior art, and does not need to increase the cost of additional hardware. It also provides strong support for automatic slag discharge adjustment of the belt conveyor when the tunneling machine is unmanned.

[0015] Further, the sparse optical flow of the target area is obtained by the following steps: two target area images at adjacent time points are obtained and preprocessed to obtain clear images, GMM algorithm and Graph Cut algorithm are used for iterative classification to segment the target area and the background in the image; Harris corner points are extracted for the target area at the current time and the last time to realize pixel calibration between the two images; a light flow field is built, and Lucas Kanade algorithm is used to obtain the sparse optical flow of the target area.

[0016] Considering that the accumulated slag is the accumulated slag on the belt conveyor and cannot be taken away by the belt conveyor, and is basically a static part in the monitoring image, the optical flow area in the monitoring image is obtained by the optical flow method, which can be considered as a moving area in the target area, and whether the accumulated slag exists is judged by the proportion of the moving area in the target area.

[0017] Further, the flow rate of the target area is calculated by the following formula: flow rate = pixel value of optical flow area / pixel value of target area.

[0018] Further, the accumulated slag treatment comprises the following steps: for the shield machine, reducing the screw machine speed to a set value and maintaining for a set time, and then judging the accumulated slag of the conveyor belt again; if the accumulated slag still exists after adjusting the screw machine speed for several times, the cutter head water spray is turned off; for the TBM, the tunneling speed is reduced and maintained for a set time, and then the accumulated slag of the conveyor belt is judged again.

[0019] For the shield machine, first, whether the accumulated slag problem can be solved by reducing the slag discharge speed is determined, and the impact on the shield process is minimized, and if the accumulated slag cannot be solved, the cutter head water spray is considered to reduce the slag humidity to solve the accumulated slag.

[0020] Further, after collecting the target area image, the following steps are further performed: filtering and gray processing of the image, then dividing the image into foreground and background according to the gray histogram, and then using the Seed Filling algorithm to obtain the maximum connected region of the foreground, and taking the proportion of the maximum connected region of the foreground as the image occlusion rate, and when the occlusion rate exceeds a set value, the water spray is used to clean the lens surface of the image acquisition device.

[0021] The image algorithm is also used to segment the foreground and background of the image, and the foreground is obviously the mud stained on the camera lens, and according to the proportion of the foreground and background, whether the camera is polluted and occluded is judged, and when the proportion is greater than a certain value, the water spray is started to solve the situation that the camera lens is polluted and occluded, which causes the method or system to malfunction, and also solves the interference on the driver caused by manual judgment of the camera occlusion.

[0022] The accumulated slag identification and monitoring system of the present application comprises a monitoring camera for collecting a target area image of a slag falling position of a conveyor belt below a slag discharge port of a tunneling machine slag discharge system and an image processing industrial computer; the image processing industrial computer calculates and obtains the sparse optical flow of the target area according to the target area image, and calculates the proportion of the optical flow in the target area as the flow rate; if the flow rate is less than a set value, it is judged that the conveyor belt is accumulated with slag.

[0023] Further, the sparse optical flow of the target region is obtained by the following steps: obtaining two target region images of adjacent time and pre-processing to obtain a clear image, using GMM algorithm and Graph Cut algorithm for iterative classification, segmenting the target region and background in the image; extracting Harris corner points for the target region of the current time and the last time, realizing pixel calibration between the two images; building an optical flow field, and using Lucas Kanade algorithm to obtain the sparse optical flow of the target region.

[0024] Further, the flow rate of the target region is calculated by the following formula: flow rate = pixel value of optical flow region / pixel value of target region.

[0025] Further, after judging that the conveying belt is accumulated with slag, the image processing industrial computer further executes the accumulated slag treatment through the heading machine upper computer; the accumulated slag treatment comprises the following steps: for the shield machine, reducing the screw machine rotating speed to a set value and maintaining for a set time, and then judging the accumulated slag of the conveying belt again; if the accumulated slag still exists after adjusting the screw machine rotating speed for several times, the cutter head water spray is turned off; for the TBM, reducing the heading speed and maintaining for a set time, and then judging the accumulated slag of the conveying belt again.

[0026] Further, after collecting the target region image, the image processing industrial computer further executes the following steps: filtering and gray processing the image, then dividing the image into foreground and background according to the gray histogram, and then using Seed Filling algorithm to obtain the maximum connected region of the foreground, and taking the proportion of the maximum connected region of the foreground as the image occlusion rate, and when the occlusion rate exceeds a set value, the water spray is used to clean the lens surface of the monitoring camera.

[0027] Compared with the prior art, the conveying belt accumulated slag identification monitoring system of the present application can solve the technical problems of accumulated slag monitoring and automatic identification, and the use of existing cameras reduces the purchase cost of related sensors, and at the same time, the sensors are not affected by slurry immersion or pollution; on the other hand, the main driver of the heading machine can pay more attention to the heading construction, and the construction safety is ensured, and technical support is provided for the automaticization of the grouting system in the subsequent unmanned heading. BRIEF DESCRIPTION OF DRAWINGS

[0028] Figure 1 is the conveying belt accumulated slag identification monitoring system of the present application;

[0029] Figure 2 is the conveying belt accumulated slag identification flowchart based on optical flow of the present application. DETAILED DESCRIPTION

[0030] In order to make the purpose, technical scheme and advantages of the present application more clear and explicit, the present application is further described in detail below with reference to the drawings and examples.

[0031] The application provides a method and system capable of accurately identifying accumulated slag of a belt conveyor system, issuing a warning and processing.

[0032] System embodiment:

[0033] This embodiment takes a shield machine as an example to illustrate the belt conveyor accumulated slag identification and monitoring system of the application. It should be understood by those skilled in the art that the accumulated slag identification system of the application can be applied to any tunneling equipment that conveys slag outward through a belt conveyor, such as a TBM. The method or system of the application should fall within the protection scope of the application when used to identify accumulated slag of various belt conveyors, including belt conveyors.

[0034] The belt conveyor accumulated slag identification and monitoring system of the application comprises a monitoring camera directed towards the position of the belt conveyor at the slag outlet of the shield machine, which is used to monitor the slag outlet and the slag on the belt conveyor, and can also determine whether the camera lens is polluted and the field of view is blocked due to the gushing or splashing of the slag through the camera image; a flushing device directed towards the camera lens from a direction that does not affect the main field of view of the camera, which is used to trigger the flushing of the camera lens when the field of view of the camera is blocked, and issue a corresponding warning; an image processing industrial computer connected to the camera, which is used to receive the camera image, run image algorithms, process and identify the camera blocking and the accumulated slag on the belt conveyor, and then transmit corresponding warning information to the shield machine host computer; a shield machine host computer connected to the image processing industrial computer, which is used to realize the adjustment of the rotation speed of the shield machine or the speed of the belt conveyor and the control of the tunneling of the shield machine, receive the warning information of the image processing industrial computer, and issue a signal instruction for flushing or operation parameter adjustment; and a network cable, which is used to connect each part to receive and transmit image information and liquid level data information.

[0035] The monitoring camera can be a camera used in the prior art to collect the image of the belt conveyor for the shield driver to observe. Alternatively, a camera that meets the requirements can be additionally installed, and the original camera can continue to be used to provide the image of the belt conveyor for the shield driver as a reference.

[0036] In the belt conveyor accumulated slag identification and monitoring system of the application, the monitoring camera is used to obtain the monitoring image shared by the slag outlet of the shield machine and the belt conveyor. After obtaining the monitoring image, the image processing industrial computer is used to realize the real-time judgment of the pollution of the camera lens by the slag due to the gushing of the shield machine, which affects the monitoring field of view. The camera blocking detection includes five steps.

[0037] ①Image preprocessing: the RGB image collected by the camera is converted into a grayscale image through grayscale processing, and Gaussian filtering is used to remove the noise in the image.

[0038] ②Division of foreground and background: the grayscale histogram of the grayscale image is solved, and the image is divided into foreground and background by adaptively finding a threshold.

[0039] ③Connected region detection, the maximum connected region of the foreground is obtained using Seed Filling algorithm;

[0040] ④Solving the occlusion rate, the pixel value of the maximum connected region of the foreground / the total pixels of the gray image is the occlusion rate η;

[0041] ⑤Issuing a warning: when η≥45%, it is considered that the image is severely occluded, an alarm message is issued, and the flushing device is started to realize the automatic flushing of the camera.

[0042] The image processing industrial computer is used to realize the recognition of the accumulated slag of the belt conveyor, as shown in FIG. Figure 2 , and specifically includes five steps:

[0043] ①Image preprocessing: first, the current monitoring image and the preset previous monitoring image are read, and then the Wiener filter is used to remove blur, ghosting and the like caused by camera shaking;

[0044] ②Grabcut target extraction: first, the target region is framed, and then the GMM algorithm and the Graph Cut algorithm are used for iterative classification to separate the target and the background in the image, and the target is below the slag outlet of the screw conveyor;

[0045] ③Detecting Harris corner points: Harris corner points are extracted for the target region of the current time and the previous time to realize the pixel calibration between the two images; ④Building an optical flow field: the target region has constant brightness, consistent space, continuous time and "small motion", and the Lucas Kanade algorithm is used to obtain the sparse optical flow of the target region;

[0046] ④Calculating the flow rate η, the flow rate is the proportion of the slurry flow in the target region, that is, the proportion of the optical flow in the target region, and the calculation formula is: flow rate = optical flow region pixel value / target region pixel value;

[0047] ⑤Issuing a warning: λ is an adjustable threshold, when the flow rate η<λ, it is identified as the accumulated slag of the belt conveyor, and a warning is issued to the main driver and the ground control center, when the flow rate η≥λ, it is identified as the normal slag discharge of the belt conveyor, and no treatment is needed. In actual tunneling, due to the poor improvement effect of the slag soil, the frequency of accumulated slag is high, after recognizing the accumulated slag and sending the warning instruction to the upper computer, the screw conveyor speed is automatically reduced to reduce the slag discharge or the cutter head water jet is reduced to treat the accumulated slag, so as to ensure that the slag soil is smoothly taken away by the belt conveyor and reduce the probability of accumulated slag.

[0048] Therefore, the belt conveyor accumulated slag monitoring system can replace manpower to realize real-time, efficient and accurate monitoring of belt conveyor accumulated slag, and realize identification, early warning and processing.

[0049] Specifically, as shown in the drawings, Figure 1 The belt conveyor accumulated slag identification and monitoring system of the present application comprises a screw conveyor 1, a screw conveyor 1 tail end is arranged downwardly to form a screw conveyor slag outlet 2; a belt conveyor 3 is arranged at the position directly below the screw conveyor slag outlet 2, and the other end of the belt conveyor 3 is directed to the direction of the tunnel outside and the direction of the slag transport vehicle; a belt conveyor and screw conveyor slag transfer area 4 is formed between the screw conveyor slag outlet 2 and the starting end of the belt conveyor 3; a monitoring camera 6 is arranged to face and cover the belt conveyor and screw conveyor slag transfer area 4, the monitoring camera 6 is fixed through a support structure 10 to ensure stable operation of the camera during shield tunneling; a flushing nozzle 5 is arranged to align with the lens of the monitoring camera 6; the monitoring camera 6 is connected through a communication network cable 9 to an image processing industrial computer 7; and the image processing industrial computer 7 is connected to a shield machine upper computer 8.

[0050] The screw conveyor 1 realizes the transfer of the slag from the soil bin to the belt, the screw conveyor slag outlet 2 realizes the discharge of the slag to the belt conveyor 3, and the high water content of the slag is easy to cause spouting and affect the field of view of the camera. The belt conveyor 3 realizes the transportation of the slag and is used to receive the slag from the screw conveyor 1 and the screw conveyor slag outlet 2. In the belt conveyor and screw conveyor outlet transfer area 4, the slag from the screw conveyor slag outlet falls to the starting end position of the belt conveyor 3 at this position by gravity, and usually this position has a certain upward angle, so when the water content of the slag is high, it is easy to cause accumulated slag. The flushing nozzle 5 is used to flush the lens of the monitoring camera 6, and the water and pressurized gas are provided by the shield machine through a pipeline. The monitoring camera 6 is used to identify the discharge situation and accumulated slag judgment; the image processing industrial computer 7 is used to store the belt conveyor accumulated slag identification algorithm model and the camera field of view blocking algorithm model, and send an early warning signal to the shield machine upper computer 8, the image processing industrial computer 7 is connected to the shield machine upper computer 8 and the monitoring camera 6 through the communication network cable 9; the shield machine upper computer 8 is used to realize the tunneling and control of the shield, and in the present application, it is used to receive the information and early warning after identification by the image processing industrial computer 7, and send a signal to adjust the tunneling parameters to solve the problem of accumulated slag, and is also used to control the flushing of the monitoring camera 6 by the flushing nozzle 5.

[0051] The tunneling machine conveyor belt accumulated slag identification and processing method of the present application is based on the above-mentioned conveyor belt accumulated slag identification and monitoring system, and when the shield machine starts tunneling, the method of the present application has two threads running simultaneously.

[0052] First thread:

[0053] (1) First determine whether the shield machine is tunneling, the specific determination method includes the following conditions: the cutter system speed in the shield machine upper computer is greater than 0.3 rad / min, the belt machine speed is >2.5 m / s, and the screw machine speed is >1.0 rad / min, where the cutter and screw machine systems are necessary components of the shield machine;

[0054] (2) When the above conditions are not met, the shield machine upper computer determines that it is not tunneling, and sends instruction 0 to the image industrial computer, representing no tunneling, at which time the work is stopped;

[0055] When the above conditions are met, the shield machine upper computer determines that it is tunneling, and sends instruction 1 to the image industrial computer, representing tunneling, at which time the monitoring camera 6 collects the monitoring image of the belt machine and screw machine slag delivery area 4, and obtains the screw machine slag delivery situation and the belt machine slag accumulation situation;

[0056] (4) After the image processing industrial computer 7 obtains the monitoring image, it filters the image to remove blur and ghosting, and then uses the Grabcut algorithm to extract the target area (below the screw machine slag delivery port), and then uses the Lucas Kanade algorithm to obtain the flow of the target area, and finally uses the flow ratio of the target area as the flow rate;

[0057] (5) After obtaining the flow rate below the screw machine slag delivery port, determine whether it exceeds the warning value, which can be specified according to the construction project site and is a fixed value, when it is higher than the warning value, at this time only provide the accumulation of slag warning information to the shield machine upper computer, when it is lower than the warning value, at this time do not provide warning information to the shield machine upper computer;

[0058] (6) After the shield machine upper computer receives the image processing industrial computer signal, if there is no warning, do not operate; if the warning is prompted, first adjust the screw machine speed to 1-2 rad / min, maintain for 20 s, continue to judge whether the slag is accumulated through the monitoring image, and feed back the result to the shield machine upper computer, if there is no effect after adjusting twice (which can be set according to needs), at this time, close the cutter spray and other actions, if there is still no effect after 20 s, stop and report.

[0059] Second thread:

[0060] (1) The image processing industrial computer 7 obtains the camera image information, first filters and processes the image to grayscale, then divides the image into foreground and background according to the grayscale histogram, and then uses the Seed Filling algorithm to obtain the maximum connected region of the foreground, and calculates the occlusion rate η of the image, when η≥45%, it is considered that the monitoring image is severely occluded.

[0061] (2) If it is determined that the monitoring image is seriously blocked, a cleaning request signal is sent to the shield machine host computer; if it is determined that it is not seriously blocked, no signal is sent to the shield machine host computer;

[0062] (3) After the shield machine host computer receives the cleaning request signal, the flushing device is started to clean the camera;

[0063] (4) If it is determined that the camera is still blocked after flushing, the flushing is continued until the flushing is clean;

[0064] The above two threads are performed simultaneously, and the determination is performed once every 1s.

[0065] Method embodiment:

[0066] The tunneling machine conveyor belt accumulated slag identification processing method is realized based on the conveyor belt accumulated slag identification monitoring system of the application. The tunneling machine conveyor belt accumulated slag identification processing method and the conveyor belt accumulated slag identification monitoring system have been sufficiently clearly described in the system embodiment, and will not be described here.

Claims

1. A method for identifying and processing slag accumulation on a tunneling machine conveyor belt, characterized in that, In the tunneling state of the tunneling machine, an image of the target area at the slag drop position of the conveyor belt below the slag outlet of the slag removal system is acquired; the GMM algorithm and Graph Cut algorithm are used to iteratively classify the current image and the previous image to segment the target area and the background in the image, and the target area is below the slag outlet of the slag removal system; Harris corner points are extracted for the target at the current time and the previous time to achieve pixel calibration between the two images; An optical flow field is constructed, and the sparse optical flow of the target area is obtained using the Lucas Kanade algorithm. The proportion of optical flow in the target area is calculated by comparing the pixel value of the optical flow area with the pixel value of the target area, which is used as the flow rate. If the flow rate is less than a set value, it is determined to be conveyor belt slag accumulation, and slag accumulation processing is performed.

2. The method for identifying and processing slag accumulation on the conveyor belt of a tunneling machine according to claim 1, characterized in that, After obtaining the image of the target area, Wiener filtering is used to remove blur and ghosting caused by camera shake.

3. The method for identifying and processing slag accumulation on the conveyor belt of a tunneling machine according to claim 1, characterized in that, The slag removal process includes the following steps: For tunnel boring machines (TBMs), reduce the screw conveyor speed to a set value and maintain it for a set time, then re-evaluate the slag accumulation on the conveyor belt; if slag accumulation persists after repeatedly adjusting the screw conveyor speed, then turn off the cutterhead water spray; For TBMs, reduce the tunneling speed and maintain it for a set time, then re-evaluate the slag accumulation on the conveyor belt.

4. The method for identifying and processing slag accumulation on the conveyor belt of a tunneling machine according to claim 1, characterized in that, After acquiring the image of the target area, the following steps are performed: the image is filtered and grayscale processed, and then the image is divided into foreground and background according to the grayscale histogram. The maximum connected region of the foreground is obtained by using the Seed Filling algorithm, and the proportion of the maximum connected region of the foreground is used as the image occlusion rate. When the occlusion rate exceeds the set value, water is sprayed to clean the surface of the lens of the image acquisition device.

5. A conveyor belt slag accumulation identification and monitoring system, characterized in that, The system includes a monitoring camera and an image processing industrial control computer for acquiring images of the target area at the slag discharge location of the conveyor belt below the slag discharge port of the tunneling machine's slag discharge system. The image processing industrial control computer performs iterative classification using the GMM algorithm and the Graph Cut algorithm based on the current image and the previous image, segmenting the target area and background in the image. The target area is below the slag discharge port of the slag discharge system. Harris corner points are extracted for the target at the current and previous times to achieve pixel calibration between the two images. An optical flow field is constructed, and the sparse optical flow of the target region is calculated using the Lucas Kanade algorithm. The proportion of optical flow in the target region is calculated by comparing the pixel value of the optical flow region with the pixel value of the target region, which is used as the flow rate. If the flow rate is less than a set value, it is judged as slag accumulation on the conveyor belt.

6. The conveyor belt slag accumulation identification and monitoring system according to claim 5, characterized in that, After obtaining the image of the target area, Wiener filtering is used to remove blur and ghosting caused by camera shake.

7. The conveyor belt slag accumulation identification and monitoring system according to claim 5, characterized in that, After determining that there is slag accumulation on the conveyor belt, the image processing industrial control computer also performs slag accumulation processing through the tunneling machine's host computer. The slag accumulation processing includes the following steps: For shield tunneling machines, after reducing the screw conveyor speed to a set value and maintaining it for a set time, the conveyor belt slag accumulation is determined again; if slag accumulation still occurs after repeatedly adjusting the screw conveyor speed, the cutterhead water spray is turned off; for TBMs, after reducing the tunneling speed and maintaining it for a set time, the conveyor belt slag accumulation is determined again.

8. The conveyor belt slag accumulation identification and monitoring system according to claim 5, characterized in that, After acquiring the image of the target area, the image processing industrial control computer also performs the following steps: filtering and grayscale processing of the image, then dividing the image into foreground and background according to the grayscale histogram, then using the Seed Filling algorithm to obtain the largest connected region of the foreground, and using the proportion of the largest connected region of the foreground as the image occlusion rate. When the occlusion rate exceeds the set value, water is sprayed to clean the surface of the monitoring camera lens.

Images